Over the past few months, the Canadian Broadcasting Corporation (CBC) launched an experiment to close comments on virtually all the news content we post to Facebook.

It was an ambitious trial. We were anxious about it. We post a lot of content to Facebook, and Facebook accounts for a healthy chunk of our referral traffic and audience engagement. Would turning off comments hurt our traffic? Would it hurt our ability to reach our audience with public service content?

We undertook the experiment because we saw an increasing amount of polarisation and hostility in the comments for our posts. There was a lot of hatred, a lot of toxicity. We wanted to reduce it and our audience’s exposure to it. We were also worried about the impact on staff that waded into these comment threads. Additionally, we found we spent a significant amount of money moderating comments on our posts.

Overall, we wanted to reduce the harm to the audience and to ourselves.

“If public discourse is a litmus test of the health of a society, the conversation on social media suggests we have a problem,” said Brodie Fenlon, editor-in-chief of CBC News. “It’s one thing for our journalists to deal with toxicity on these platforms. It’s another for our audience members who try to engage with and discuss our journalism to encounter it on platforms such as Facebook and Twitter, where they are almost guaranteed to be confronted by hate, racism, and abuse.” 

At the time we launched the experiment, we didn’t know specifically that Facebook’s algorithm helped to foster discord, as The Wall Street Journal recently reported. However, we had a sense the comments had been getting worse over the years.

So, we went ahead with the experiment. We turned off all the comments on all news story links and native video posted to our Facebook pages.

“This is a proud moment for me and my association with CBC. We are tackling a big problem from a learning posture,” said Richard Kanee, the head of our digital products department, at the time. “We are challenging assumptions deeply embedded in the publishing industry. We are taking steps to address the erosion of civil discourse.”

The experiment started in mid-June and ends in October.

What was the impact?

The results have surprised us in a few different ways.

Our staff members realised they could now post a broader range of content that previously would have attracted hate, misogyny, and death threats in the comments of our Facebook posts. Our Facebook pages now circulated more stories with a diversity of voices, experiences, and communities. This allowed us to better fulfill our public service mandate of informing, enlightening, and reflecting all people of Canada.

At the same time, our staff told us they still want to be able to engage with Facebook audiences in some way. For the last part of this experiment, we’ve set out specific scenarios to try opening comments on posts aimed at soliciting people’s experiences and reactions that will help enrich our journalism, such as identifying interview sources or follow-up story ideas.

So far, we’ve found that turning off comments on our Facebook posts has had little effect on the volume of referrals we see from Facebook. “It’s pretty clear that this experiment has not had a material negative impact on overall traffic,” said Andy Baker, the research lead on this experiment.

We’ve not seen a big dip in referral traffic from Facebook, nor have we seen a large decrease in the amount of time audiences from Facebook spend on our site. This is likely due to the fact that most of the referral traffic from Facebook doesn’t come from our posts, but from what the audience is posting; it’s traffic from organic sharing.

We can conclude that the traffic from organic sharing is larger than what we expected, and it accounts for a larger proportion of Facebook referrals than we thought. It is also possible that Facebook treats this kind of content differently, and closing comments on this content doesn’t impact reach on the platform. We don’t have access to the data to know for sure.

How did the audience react?

When we announced the experiment in an editor’s blog post, we got more than 8,000 comments, which is well above average. The experiment clearly struck a nerve with the audience.

Some of the commenters were in favour of it. Some were against it. One audience member said: “I am absolutely disgusted that you are blocking comments. CBC used to stand for community and FREE speech. It is the beginning of the end of our beautiful country.”

On the other hand, others were more supportive: “I am so pleased to read that you are protecting your staff from what can be a very toxic environment at this time. Well done. As a former school principal, I am very aware of the negative impact social media platforms can have when used by some to vent and rant. Bravo.”

Another person wrote to us, “Now that comments are closed, I just read the news and it feels much healthier.”

We’ve also heard from the public about how closing comments has helped reduce the spread of disinformation. One reader shared the following: “I have elderly family members who have been led down a dangerous path of conspiracy and misinformation from CBC comments on the Web site and Facebook — they tell me they see info on ‘CBC’ and every time I check, it’s in the comments even on Facebook.”

Overall, the feedback from the audience was about evenly split. On our feedback channels, about 50% agreed with the move, while about 50% were opposed to it.

We also observed a small number of people intent on spreading their vitriol have migrated to Facebook Messenger to direct their toxic messages to our staff.

What’s next?

First, we’ll wait until the experiment finishes at the end of October, and then we’ll evaluate the final results. If the results that we’ve seen so far don’t change, we expect this experiment will teach us a lot about the value of opening or closing comments on Facebook. We’ve already learned that closing comments on Facebook has not resulted in a large impact on referrals or engagement from Facebook.

We will also be asking some tough questions about the appropriate degree to which the CBC should be participating on the platform, especially given some of the revelations from The Wall Street Journal coverage in the last couple of weeks.

“We have to continue to be bold and intentional about how and when we use Facebook to create value for our audience and our journalism,” said Andree Lau, managing editor of digital news at CBC and one of the experiment leads. “This test is generating valuable data and observations to guide our decisions on sustainable and healthy ways to serve the Canadian public and fulfill our mandate.”

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .